This article is more than 5 years old. Information may no longer be current.
One in four coded TJA readmission diagnoses did not match clinical diagnosis: study
Click Here to Manage Email Alerts
A study that researchers did into coded diagnoses compared to clinician diagnoses for patients who were readmitted after total joint arthroplasty found agreement occurred 75% of the time, according to this study.
“Our manual review revealed that the language of the medical chart was often unclear. For example, for patients readmitted with suspected [deep vein thrombosis] DVT, the language sometimes remained vague throughout the encounter (e.g., “patient admitted with likely DVT” or more generically “patient admitted with lower extremity swelling”), even in the presence of positive or negative diagnostic testing,” S. David Stulberg, MD, and colleagues at the Northwestern University Feinberg School of Medicine, in Chicago, wrote in their study.
S. David Stulberg
“Working with our clinical coding department, we learned that the coding process essentially represents a retrospective review of each hospital encounter that does not actually begin until discharge,” Stulberg and colleagues noted. “Moreover, those responsible for coding are restricted to the data available in clinical documentation performed by physicians and mid-level providers; they are not permitted to use lab or imaging results.”
The researchers studied 87 patients who were readmitted 90 days after surgery and found the most common reasons for readmission were procedure-related, according to the abstract. There was a 7.9% overall readmission rate for joint arthroplasty procedures done at the institution.
For the patients included in the study, the coded diagnosis for 22 patients did not match the clinician diagnosis, according to the abstract. Coding discrepancies occurred in cases of postoperative stiffness and when manipulation was needed after total knee arthroplasty, as well as when other implant, prosthetic and graft-based complications occurred, Stulberg and colleagues noted.
|
|
In the discussion of their results, Stulberg and colleagues expressed concern about using readmissions data for reimbursement by the CMS based on their study’s data.
“Direct physician entry of diagnoses through the electronic medical record may improve the accuracy of principal diagnoses, but hospital coders will still depend on appropriate documentation of secondary conditions,” Stulberg and colleagues wrote. “Given the expectation that ICD-10 is potentially more granular and thus more cumbersome than our current system, such inconsistency may become more prevalent.”
Disclosure: Stulberg receives royalties from Aesculap and Innomed, is on the speaker’s bureau for Zimmer and is a paid consultant for Aesculap, Stryker, Zimmer, OmniLife Sciences and Innomed. Puri is a paid consultant for Stryker and Salient.
Click Here to Manage Email Alerts